chore(notice): pin exo provenance and deviations; keep GPL-3.0; sweep product branding

NOTICE

@@ -7,6 +7,11 @@ Public License, version 3 (GPL-3.0):
 
     Copyright (c) 2024 exo labs (Exo Technologies Ltd.)
 
+Forked from https://github.com/exo-explore/exo (GPL-3.0). The exact upstream
+commit was not recorded at fork time; this tree was taken from exo's GPL-3.0
+line (package version 0.1.x) prior to the exo v1.0 (Apache-2.0) relicense of
+December 2025.
+
 Hanzo Net is a fork of exo. It was forked while exo was licensed under
 GPL-3.0, and Hanzo Net therefore REMAINS licensed under GPL-3.0 (see the
 LICENSE file). Hanzo AI has NOT relicensed this work.
@@ -25,6 +30,24 @@ generation and mobile WebGPU onboarding; host/port configuration; and
 graceful-shutdown handling. See the git history for the full, dated record
 of changes.
 
+DEVIATIONS (per GPL-3.0 section 5 — this fork differs from upstream exo as
+follows; see the git history for the dated, per-commit record):
+
+  - Renamed the upstream `exo` Python package to `net`; import path is
+    `net.*` and the console entry point is `hanzo-net`.
+  - Rebranded the product to "Hanzo Net" and the web UI to "Hanzo Cluster";
+    removed the upstream "tinychat" label from the Web Chat URL display.
+  - Published as the `hanzo-net` distribution under Hanzo AI, Inc.
+    (author dev@hanzo.ai, repository github.com/hanzoai/net).
+  - Responsive CLI rendering that adapts to terminal width for small screens.
+  - Real QR-code generation for onboarding that encodes the actual LAN IP
+    (not localhost) so phones and tablets can join the cluster.
+  - Mobile onboarding path with WebGPU support via the web UI.
+  - Configurable host and port for the node and its API.
+  - Graceful shutdown on Ctrl-C that suppresses spurious gRPC teardown errors.
+  - Added OSS attribution: this NOTICE plus retained upstream copyright
+    headers in vendored files.
+
 --------------------------------------------------------------------------
 
 This product also includes model code from MLX examples

setup.py

@@ -91,7 +91,7 @@ _add_gpu_requires()
 setup(
   name="hanzo-net",
   version="0.1.20",
-  description="Hanzo Network - Distributed AI compute network for running models locally and remotely",
+  description="Hanzo Net - Distributed AI compute network for running models locally and remotely",
   author="Hanzo AI",
   author_email="dev@hanzo.ai",
   url="https://github.com/hanzoai/net",
@@ -112,7 +112,7 @@ setup(
     "Operating System :: OS Independent",
   ],
   long_description="""
-Hanzo Network - Distributed AI compute network for running models locally and remotely.
+Hanzo Net - Distributed AI compute network for running models locally and remotely.
 
 **Platform Support:**
 - Linux: Full support